Privacy and data protection Archives | Page 108 of 349

Growing app restrictions hit ByteDance’s AI smartphone rollout

ByteDance is facing mounting pushback after major Chinese apps restricted how its agentic AI smartphone can operate across their platforms. Developers moved to block or limit Doubao, the device’s voice-driven assistant, following concerns about automation, security and transactional risks.

Growing reports from early adopters describe locked accounts, interrupted payments and app instability when Doubao performs actions autonomously. ByteDance has responded by disabling the assistant’s access to financial services, rewards features and competitive games while collaborating with app providers to establish clearer guidelines.

The Nubia M153, marketed as an experimental device, continues to attract interest for its hands-free interface, even as privacy worries persist over its device-wide memory system. Its long-term success hinges on whether China’s platforms and regulators can align with ByteDance’s ambitions for seamless, AI-powered smartphone interaction.

Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!

New interview study tracks how workers adapt to AI

Anthropic has unveiled Anthropic Interviewer, an AI-driven tool for large-scale workplace interviews. The system used Claude to conduct 1,250 structured interviews with professionals across the general workforce, creative fields and scientific research.

In surveys, 86 percent said AI saves time and 65 percent felt satisfied with its role at work. Workers often hoped to automate routine tasks while preserving responsibilities that define their professional identity.

Creative workers reported major time savings and quality gains yet faced stigma and economic anxiety around AI use. Many hid AI tools from colleagues, feared market saturation and still insisted on retaining creative control.

Across groups, professionals imagined careers where humans oversee AI systems rather than perform every task themselves. Anthropic plans to keep using Anthropic Interviewer to track attitudes and inform future model design.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Japan weighs easing rules on personal data use

Japan is preparing to relax restrictions on personal data use to support rapid AI development. Government sources say a draft bill aims to expand third-party access to sensitive information.

Plans include allowing medical histories and criminal records to be obtained without consent for statistical purposes. Japanese officials argue such access could accelerate research while strengthening domestic competitiveness.

New administrative fines would target companies that profit from unlawfully acquired data affecting large groups. Penalties would match any gains made through misconduct, reflecting growing concern over privacy abuses.

A government panel has reviewed the law since 2023 and intends to present reforms soon. Debate is expected to intensify as critics warn of increased risks to individual rights if support for AI development in this regard continues.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Australia seals $4.6 billion deal for new AI hub

OpenAI has partnered with Australian data centre operator NextDC to build a major AI campus in western Sydney. The companies signed an agreement covering development, planning and long-term operation of the vast site.

NextDC said the project will include a supercluster of graphics processors to support advanced AI workloads. Both firms intend to create infrastructure capable of meeting rapid global demand for high-performance computing.

Australia estimates the development at A$7 billion and forecasts thousands of jobs during construction and ongoing roles across engineering and operations. Officials say the initiative aligns with national efforts to strengthen technological capability.

Plans feature renewable energy procurement and cooling systems that avoid drinking water use, addressing sustainability concerns. Treasurer Jim Chalmers said the project reflects growing confidence in Australia’s talent, clean energy capacity and emerging AI economy.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Noyb study points to demand for tracking-free option

A new study commissioned by noyb reports that most users favour a tracking-free advertising option when navigating Pay or Okay systems. Researchers found low genuine support for data collection when participants were asked without pressure.

Consent rates rose sharply when users were presented only with payment or agreement to tracking, leading most to select consent. Findings indicate that the absence of a realistic alternative shapes outcomes more than actual preference.

Introduction of a third option featuring advertising without tracking prompted a strong shift, with most participants choosing that route. Evidence suggests users accept ad-funded models provided their behavioural data remains untouched.

Researchers observed similar patterns on social networks, news sites and other platforms, undermining claims that certain sectors require special treatment. Debate continues as regulators assess whether Pay or Okay complies with EU data protection rules such as the GDPR.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

NITDA warns of prompt injection risks in ChatGPT models

Nigeria’s National Information Technology Development Agency (NITDA) has issued an urgent advisory on security weaknesses in OpenAI’s ChatGPT models. The agency warned that flaws affecting GPT-4o and GPT-5 could expose users to data leakage through indirect prompt injection.

According to NITDA’s Computer Emergency Readiness and Response Team, seven critical flaws were identified that allow hidden instructions to be embedded in web content. Malicious prompts can be triggered during routine browsing, search or summarisation without user interaction.

The advisory warned that attackers can bypass safety filters, exploit rendering bugs and manipulate conversation context. Some techniques allow injected instructions to persist across future interactions by interfering with the models’ memory functions.

While OpenAI has addressed parts of the issue, NITDA said large language models still struggle to reliably distinguish malicious data from legitimate input. Risks include unintended actions, information leakage and long-term behavioural influence.

NITDA urged users and organisations in Nigeria to apply updates promptly and limit browsing or memory features when not required. The agency said that exposing AI systems to external tools increases their attack surface and demands stronger safeguards.

Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!

EU gains stronger ad oversight after TikTok agreement

Regulators in the EU have accepted binding commitments from TikTok aimed at improving advertising transparency under the Digital Services Act.

An agreement that follows months of scrutiny and addresses concerns raised in the Commission’s preliminary findings earlier in the year.

TikTok will now provide complete versions of advertisements exactly as they appear in user feeds, along with associated URLs, targeting criteria and aggregated demographic data.

Researchers will gain clearer insight into how advertisers reach users, rather than relying on partial or delayed information. The platform has also agreed to refresh its advertising repository within 24 hours.

Further improvements include new search functions and filters that make it easier for the public, civil society and regulators to examine advertising content.

These changes are intended to support efforts to detect scams, identify harmful products and analyse coordinated influence operations, especially around elections.

TikTok must implement its commitments to the EU within deadlines ranging from two to twelve months, depending on each measure.

The Commission will closely monitor compliance while continuing broader investigations into algorithmic design, protection of minors, data access and risks connected to elections and civic discourse.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

EU ministers call for faster action on digital goals

European ministers have adopted conclusions aimed to boosting the Union’s digital competitiveness, urging quicker progress toward the 2030 digital decade goals.

Officials called for stronger digital skills, wider adoption of technology, and a framework that supports innovation while protecting fundamental rights. Digital sovereignty remains a central objective, framed as open, risk-based and aligned with European values.

Ministers supported simplifying digital rules for businesses, particularly SMEs and start-ups, which face complex administrative demands. A predictable legal environment, less reporting duplication and more explicit rules were seen as essential for competitiveness.

Governments emphasised that simplification must not weaken data protection or other core safeguards.

Concerns over online safety and illegal content were a prominent feature in discussions on enforcing the Digital Services Act. Ministers highlighted the presence of harmful content and unsafe products on major marketplaces, calling for stronger coordination and consistent enforcement across member states.

Ensuring full compliance with EU consumer protection and product safety rules was described as a priority.

Cyber-resilience was a key focus as ministers discussed the increasing impact of cyberattacks on citizens and the economy. Calls for stronger defences grew as digital transformation accelerated, with several states sharing updates on national and cross-border initiatives.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Australia introduces new codes to protect children online

Australian regulators have released new guidance ahead of the introduction of industry codes designed to protect children from exposure to harmful online material.

The Age Restricted Material Codes will apply to a wide range of online services, including app stores, social platforms, equipment providers, pornography sites and generative AI services, with the first tranche beginning on 27 December.

The rules require search engines to blur image results involving pornography or extreme violence to reduce accidental exposure among young users.

Search services must also redirect people seeking information related to suicide, self-harm or eating disorders to professional mental health support instead of allowing harmful spirals to unfold.

eSafety argues that many children unintentionally encounter disturbing material at very young ages, often through search results that act as gateways rather than deliberate choices.

The guidance emphasises that adults will still be able to access unblurred material by clicking through, and there is no requirement for Australians to log in or identify themselves before searching.

eSafety maintains that the priority lies in shielding children from images and videos they cannot cognitively process or forget once they have seen them.

These codes will operate alongside existing standards that tackle unlawful content and will complement new minimum age requirements for social media, which are set to begin in mid-December.

Authorities in Australia consider the reforms essential for reducing preventable harm and guiding vulnerable users towards appropriate support services.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

EU targets X for breaking the Digital Services Act

European regulators have imposed a fine of one hundred and twenty million euros on X after ruling that the platform breached transparency rules under the Digital Services Act.

The Commission concluded that the company misled users with its blue checkmark system, restricted research access and operated an inadequate advertising repository.

Officials found that paid verification on X encouraged users to believe their accounts had been authenticated when, in fact, no meaningful checks were conducted.

EU regulators argued that such practices increased exposure to scams and impersonation fraud, rather than supporting trust in online communication.

The Commission also stated that the platform’s advertising repository lacked essential information and created barriers that prevented researchers and civil society from examining potential threats.

European authorities judged that X failed to offer legitimate access to public data for eligible researchers. Terms of service blocked independent data collection, including scraping, while the company’s internal processes created further obstacles.

Regulators believe such restrictions frustrate efforts to study misinformation, influence campaigns and other systemic risks within the EU.

X must now outline the steps it will take to end the blue checkmark infringement within sixty working days and deliver a wider action plan on data access and advertising transparency within ninety days.

Failure to comply could lead to further penalties as the Commission continues its broader investigation into information manipulation and illegal content across the platform.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!